Nuclear Technology / Volume 156 / Number 3 / December 2006 / Pages 289-302
Technical Paper / Thermal Hydraulics / dx.doi.org/10.13182/NT06-A3792
In a thermal-hydraulic analysis for nuclear application, a one-dimensional analysis is widely used. In the analysis, averaging is required for the calculation of the cell property, and the accuracy of the averaging directly influences the accuracy of a numerical scheme. While the average value depends on the property distribution characteristics in a cell, conventional numerical schemes do not utilize the information. Instead, they rely on the use of a large number of nodes for their accuracy. There are many cases where the use of a large number of nodes is not practically allowed, especially in a transient system analysis, and the calculation results come to suffer from a large truncation error. To overcome the drawbacks of the conventional schemes, a new approach is introduced to reduce the truncation error by utilizing the distribution characteristics in a cell for the required averaging. The new approach places a node point at the boundary of a calculation cell and averaging is achieved from the properties at the inlet and outlet by using weighting factors that are determined from the cell property distribution. By this approach, it was successful to describe more accurately even a transient where the property distribution was stepwise. Steady-state calculation for a once-through steam generator where the feedwater is heated to superheated steam was accurately carried out with only three calculational nodes. The characteristics and achievements of the new approach are described.